In traditional time series forecasting problems, researchers mostly used classical time series models, such as, Autoregressive (AR), Moving average (MA) and ARIMA to model linear time series data. On the other hand, ...